Written by Yi-An Chen. Image credit to News Center.
Prof. Trevor Darrell of University of California, Berkeley was invited to NCKU to share his latest research outcome and insight about AI development on January 11st, 2023.
Trevor Darrell, the Professor of the CS and EE Divisions of the EECS Department at UC Berkeley.
Prof. Trevor Darrell is the founding Co-Director of Berkeley Artificial Intelligence Research (BAIR), Berkeley Deep Drive (BDD), and BAIR Commons. In his speech “Unsupervised, Adaptive, and Advisable Visual Learning”, he mentioned that unsupervised and cross-modal methods are rapidly overtaking supervised techniques as the dominant paradigm for visual learning, and models based on reconstruction, masking, and denoising have demonstrated ability to generalize from few examples.
Prof. Darrell noted that with the combination of language and gesture, AI can expand more deep-learning ability than simply rely on visual image.
Prof. Darrell shared his recent advances at Berkeley developing such models for visual pretraining of robotic agents, learning visual analogies, adapting to out-of-distribution data, and learning social gestures, and will speculate on future unified models as time permits. Plenty NCKU students were attracted to participate in Prof. Darrell’s speech and even stay for more discussion with lecturer after the speech ended.
Prof. Darrell brought cutting-edge insight of AI research to NCKU students. People gathered around Prof. Darrell for more discussion.
Prof. Darrell said it was his pleasure to be invited to Taiwan and share his opinion with Taiwan scholars. His team is not only working on expanding the application fields of deep learning, but also include the topics of computer vision, machine learning, natural language processing, and perception-based human computer interfaces.
Prof. Pau-Choo Jang, the Dean of the Miin Wu School of Computing stated that Prof. Darrell is more than a Professor of the CS and EE Divisions of the EECS Department at UC Berkeley, he is also the co-founder of BAIR Commons program, a partner company of well-known companies, such as Facebook, Google, Microsoft and Amazon. The perceptual learning program develop by their researching team can apply to self-driving cars, multimedia research and multimodal interaction between machine and human beings.
Dr. Ching-Yao Chan was praised to be the Taiwanese that know best of self-driving cars in the world.
Dr. Ching-Yao Chan, the co-director of Berkley DeepDrive (BDD), a research consortium focused on computer vision and AI technologies, visited NCKU with Prof. Darrell January 11st, 2023 as well. Dr. Chan has dive into the researching area of self-driving cars for more than 10 years and was praised to be the Taiwanese that know best of self-driving cars in the world. Dr. Chan is now serving as a member of board of Miin Wu School of Computing of NCKU.
Dr. Ching-Yao Chan mentioned that the AI technology with cross-media capability is a new trend and a crucial issue for related researching field. He also said that people used to apply AI to image analyzation only, but Prof. Darrell utilized it with more aspect, such as gestures, languages, visual images, social media and so on, thus creating a model of AI and deep-learning, which is important research development that make AI closer to real life.